检索结果-内蒙古大学图书馆

Computing for Sustainability and Intelligent Future (COMP-SIF), International Conference on

作者： Ravikumar J Prithvi Prabhu Pani V Tejushree R Sunnampalli Varshitha Department of Computer Science and Engineering Atria Institute of Technology Bangalore India

ISBN: (数字)9798331538538

ISBN: (纸本)9798331538545

Breast cancer is one of the leading causes of death in women. Regular testing is of paramount importance since early detection ensures more treatment options and better prognosis rate. Biopsies, x-rays and ultrasounds are some of the diagnostic tools used to detect breast cancer. Out of these, ultrasounds being portable and ionizing radiation free provide the most feasible and harmless solution in areas which cannot accommodate expensive equipment that require rigorous safety standards to be met. Our detection system utilizes Vision Transformer (ViT) technology, i.e., transfer learning to classify tumors of the breast into malignant, benign and normal categories. Data enhancement has been done using Contrast Limited Adaptive Histogram Equalization (CLAHE). Furthermore, we have categorized them based on BI-RADS criteria for greater accuracy and precision scores. We have chosen ViT over the more commonly used Convolution Neural Networks (CNN) since they offer a more flexible approach, capturing global relationships in images, and you can use it for larger datasets.

关键词： computer vision Ultrasonic imaging Accuracy Transfer learning Imaging X-rays Transformers Breast cancer Convolutional neural networks Tumors

来源：评论

学校读者我要写书评

暂无评论

Real-time Air Quality Monitoring and Context-aware Alert System

Real-time Air Quality Monitoring and Context-aware Alert Sys...

引用

International Communication Systems and Networks and Workshops, COMSNETS

作者： Anuj Kakde Yashwant Krishna Pagoti Sarthak Nikumbh Prasenjit Karmakar Sandip Chakraborty Department of Computer Science and Engineering Indian Institute of Technology Kharagpur India

ISBN: (数字)9798331531195

ISBN: (纸本)9798331531201

Air quality monitoring is gaining increasing importance as awareness about the health impacts of air pollution continues to grow. These monitors typically track air pollutants to give users a clearer picture of their environment. Accurate and real-time monitoring increases our understanding of the surrounding air we breathe and allows us to improve it. The proposed system helps users minimize their exposure to harmful pollutants by providing real-time data on the concentration of various gases. The system triggers alerts directly to the subscribed user’s mobile phones in case of any hazardous situations so that the user can take appropriate actions immediately to resolve the issue. Our extensive data collection and evaluation with off-the-shelf machine learning models indicated an overall F1-score of 99.54 (±0.5)% with a simple random forest model in classifying burning, cooking, occupancy in a closed room, using aerosol products, based on the sensor’s readings.

关键词： Atmospheric modeling Aerosols Air pollution Ventilation Real-time systems Data models Hazards Monitoring Random forests Smart phones

来源：评论

学校读者我要写书评

暂无评论

Leveraging EfficientNet-B0 for Soybean Leaf Disease Detection: A Deep Learning Perspective

Leveraging EfficientNet-B0 for Soybean Leaf Disease Detectio...

引用

Computational, Communication and Information Technology (ICCCIT), International Conference on

作者： Sanket Gupta Suresh Jain Kailash Chandra Bandhu Department of Computer Science Engineering Medi-Caps University Indore India

ISBN: (数字)9798331512965

ISBN: (纸本)9798331512972

Plant diseases require quick and accurate identification methods since they significantly reduce agricultural production's output and quality. Traditional methods of plant protection, which depend on visual assessment of pests and diseases that impact soyabean plants, have long turnaround times and variable accuracy. A critical need as the demand for precision agriculture grows is the creation of quick, effective, and—above all—computer-aided disease recognition systems. This study proposed convolutional neural network-based classification model for soyabean leaf. Proposed model can detect nine different types of soyabean diseases. Our transfer-learning based model classifies soybean leaves using a Inception B0 convolutional neural network, and on unseen testing data, we obtain balanced accuracies between 91.11% and 96.44%. The result of proposed model is generated on 3000 different images of soyabean leaves and classify nine different types of soyabean diseases.

关键词： Deep learning Precision agriculture Visualization Plant diseases Accuracy Transfer learning Production Convolutional neural networks Protection Testing

来源：评论

学校读者我要写书评

暂无评论

At the Mahakumbh, Faith Met Tragedy: Computational Analysis of Stampede Patterns Using Machine Learning and NLP

arXiv

引用

arXiv 2025年

作者： Pratap, Abhinav Department of Computer Science and Engineering ASET Amity University Noida India

This study employs machine learning, historical analysis, and natural language processing (NLP) to deconstruct the recurring lethal stampedes at India’s mass religious gatherings, focusing on the 2025 Mahakumbh tragedy in Prayagraj (48+ deaths) and its 1954 predecessor (700+ casualties). Through comparative computational modeling of crowd dynamics and administrative records, it examines how systemic vulnerabilities transmute devotion into disaster. Temporal trend analysis identifies persistent choke points, with narrow riverbank access routes linked to 92% of past stampede sites and lethal crowd densities (≥8 persons/m2) recurring during spiritually charged moments like Mauni Amavasya [1][2]. NLP analysis of seven decades of inquiry reports reveals cyclical administrative failures, where VIP route prioritization diverted safety resources in both 1954 and 2025, exacerbating fatalities [3]. Statistical modeling demonstrates how ritual urgency overrides risk perception, leading to panic propagation patterns that mirror historical incidents [4]. Findings support Institutional Amnesia Theory, highlighting how disaster responses remain reactionary rather than preventive [5]. By correlating archival patterns with computational crowd behavior analysis, this study frames stampedes as a collision of infrastructure limitations, socio-spiritual urgency, and governance inertia, challenging disaster discourse to confront how spiritual economies normalize preventable mortality. © 2025, CC BY.

关键词： Risk perception

来源：评论

学校读者我要写书评

暂无评论

End-to-end Training for Text-to-Image Synthesis using Dual-Text Embeddings

arXiv

引用

arXiv 2025年

作者： Ahmed, Yeruru Asrar Mittal, Anurag Department of Computer Science and Engineering Indian Institute of Technology Madras India

Text-to-Image (T2I) synthesis is a challenging task that requires modeling complex interactions between two modalities (, i.e., text and image). A common framework adopted in recent state-of-the-art approaches to achieving such multimodal interactions is to bootstrap the learning process with pre-trained image-aligned text embeddings trained using contrastive loss. Furthermore, these embeddings are typically trained generically and reused across various synthesis models. In contrast, we explore an approach to learning text embeddings specifically tailored to the T2I synthesis network, trained in an end-to-end fashion. Further, we combine generative and contrastive training and use two embeddings, one optimized to enhance the photo-realism of the generated images, and the other seeking to capture text-to-image alignment. A comprehensive set of experiments on three text-to-image benchmark datasets (Oxford-102, Caltech-UCSD, and MS-COCO) reveal that having two separate embeddings gives better results than using a shared one and that such an approach performs favourably in comparison with methods that use text representations from a pre-trained text encoder trained using a discriminative approach. Finally, we demonstrate that such learned embeddings can be used in other contexts as well, such as text-to-image manipulation. © 2025, CC BY.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Privacy-Preserving Multimodal Sentiment Analysis

引用

IEEE Internet of Things Journal 2025年第11期12卷 15467-15478页

作者： Xu, Honghui Li, Wei Takabi, Daniel Seo, Daehee Cai, Zhipeng Kennesaw State University Department of Information Technology KennesawGA United States Georgia State University Department of Computer Science AtlantaGA United States Old Dominion University School of Cybersecurity NorfolkVA United States Sangmyung University College of Intelligence Information Engineering Seoul Korea Republic of

Multimodal sentiment analysis plays a critical role in numerous IoT-driven applications, such as personalized smart assistants, healthcare monitoring systems, and intelligent transportation networks, where accurate interpretation of user emotions is vital for enhancing service quality. However, a severe threat of privacy leakage in the multimodal sentiment analysis has been overlooked by previous works. To fill this gap, we propose a Differentially Private Correlated Representation Learning (DPCRL) model to achieve privacy-preserving multimodal sentiment analysis by combining a correlated representation learning scheme with a differential privacy protection scheme. Our correlated representation learning scheme aims to achieve heterogeneous multimodal data transformation to meet the requirements of privacy-preserving multimodal sentiment analysis by learning the correlated and uncorrelated representations, where especially, a pre-determined correlation factor is employed to flexibly adjust the expected correlation among the correlated representations. The differential privacy protection scheme is used to obtain the disturbed correlated and uncorrelated representations by adding Laplace noise for -differential privacy. In particular, the correlation factor can help alleviate the side-effect of the added Laplace noise on the sentiment prediction performance. Finally, via conducting a series of real-data experiments, we validate that our proposed DPCRL model is superior to the state of the art for privacy-preserving multimodal sentiment analysis. © 2014 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

AI-Enhanced Pod Scheduling: Optimizing MERN and MEAN Stack Performance in Kubernetes

AI-Enhanced Pod Scheduling: Optimizing MERN and MEAN Stack P...

引用

Data science, Agents & Artificial Intelligence (ICDSAAI), International Conference on

作者： Nivetha Ramachandran M. Thirumaran Department of Computer Science and Engineering Puducherry Technological University Puducherry India

ISBN: (数字)9798331537555

ISBN: (纸本)9798331537562

Microservices architecture has become the foundation of modern application development, offering enhanced scalability, flexibility, and maintainability. This paper presents a comparative performance evaluation of two widely used JavaScript-based stacks, MERN (MongoDB, Express, React, ***) and MEAN (MongoDB, Express, Angular, ***), deployed as microservices applications in Minikube, a local Kubernetes environment. While both stacks share a common backend technology stack, they differ in their front-end frameworks—React in MERN and Angular in MEAN—which influences their performance, scalability, and suitability for different application types. This study not only compares the architectural differences between MERN and MEAN but also introduces machine learning (ML)-based microservice optimization to enhance service instance selection and resource allocation in Kubernetes. The optimization leverages RF (Random Forest) and XGBoost ML algorithms for dynamic intelligent POD selection and scheduling, ensuring dynamic routing of microservice requests based on real-time performance metrics. Key evaluation criteria include latency, response time, CPU and memory utilization, scalability, and fault tolerance, monitored using Prometheus and Istio. The results provide insights into the efficiency of each stack in handling microservices workloads, helping developers and system architects choose the most suitable stack based on performance demands, resource constraints, and deployment complexity. The paper concludes with recommendations for optimizing microservices-based applications in Kubernetes environments using ML-driven orchestration strategies.

关键词： Radio frequency Heuristic algorithms Scalability Microservice architectures Aerodynamics Dynamic scheduling Time factors Resource management Random forests Monitoring

来源：评论

学校读者我要写书评

暂无评论

Comparative Analysis of Advanced Blockchain Authentication Model for Loan and Tender Sanctioning

Comparative Analysis of Advanced Blockchain Authentication M...

引用

Advancements in Electrical, Electronics, Communication, Computing and Automation (ICAECA), International Conference on

作者： Saranyah V N Suganthi Department of Computer Science and Engineering Kumaraguru College of Technology Coimbatore India

ISBN: (数字)9798331525439

ISBN: (纸本)9798331525446

The conventional method of tender sanctioning involves a process whereby tenders are awarded to companies that offer the most efficient and economical solutions according to the requirements of the government or the providing organization. However, the selection criteria for these companies are often shrouded in secrecy, lacking transparency in the allocation of tenders to specific bidding parties. To address this issue, we propose a blockchain-based authentication scheme incorporating different consensus mechanisms at various stages of the tender sanctioning process. This approach ensures a secure and transparent method of tender sanctioning. Previously we have implemented our concept in another aspect of bank operations, specifically in the sanctioning of loans to eligible customers of the bank. By utilizing our proposed method, we establish a transparent process for loan sanctioning within the bank. Furthermore, we conduct a comparative analysis to evaluate the effectiveness of our proposed method across two distinct applications. This analysis involves comparing the efficacy of various consensus mechanisms, such as Proof of Work and Proof of Authority, in both scenarios.

关键词： Procurement Scalability Authentication Banking Companies Proof of Work Consensus protocol Telecommunication computing Security Resource management

来源：评论

学校读者我要写书评

暂无评论

Exploring Mental Well-Being Based on Daily Routine Patterns and on the Impact of Sleep Quality

Exploring Mental Well-Being Based on Daily Routine Patterns ...

引用

International Conference on Electrical, computer and Communication engineering (ECCE)

作者： Md. Rubayed Hossain Sumaiya Akter Ritu Md. Tauhid Md. Firoz Hasan Department of Computer Science & Engineering Daffodil International University Dhaka Bangladesh

ISBN: (数字)9798350357509

ISBN: (纸本)9798350357516

This research examines the influence of regularity of routine and quality of sleep on mental well-being, providing evidence-based solutions for these basic relationships. From the data of 376 participants, the strongest predictors of mental well-being were determined to be regularity of routine and ideal duration of sleep ranging from 7–9 hours. Mental wellbeing states were classified using machine learning techniques, and XGBoost provided an accuracy of 88.20%. The results offer practical advice on how to improve mental functioning by situating healthy practices and adequate rest at the core as key drivers for building mental toughness.

关键词： Accuracy Computational modeling Buildings Mental health Predictive models Robustness Distance measurement Random forests Optimization

来源：评论

学校读者我要写书评

暂无评论

DQN-Based Crypto Trading: A Hybrid Approach to Market-Aware Buy-Hold Decisions and Profit-Driven Selling

DQN-Based Crypto Trading: A Hybrid Approach to Market-Aware ...

引用

Advancements in Electrical, Electronics, Communication, Computing and Automation (ICAECA), International Conference on

作者： C. Bharathi Priya Murugesan V Sivavishnu G Vignesh Meenakshi Sundaram Department of Computer Science & Engineering Kumaraguru College of Technology Coimbatore India

ISBN: (数字)9798331525439

ISBN: (纸本)9798331525446

Cryptocurrencies have quickly become a significant part of the financial landscape, largely due to their potential for substantial returns. This study examines a trading model based on Deep Q-Network (DQN) that combines reinforcement learning with a hybrid decision-making approach. Unlike traditional RL models that treat buy, hold, and sell decisions as equal, this framework differentiates the decision-making process. It enables DQN to learn optimal strategies for buying and holding while using a rule-based mechanism focused on profit targets for selling. This hybrid method ensures the model effectively captures market trends and secures profits at set thresholds, thereby minimizing unnecessary risk exposure. By adapting to changing market conditions and automating trade execution, this approach seeks to enhance profitability while reducing the risks associated with cryptocurrency trading. Experimental results highlight the effectiveness of this method in volatile market conditions

关键词： Deep learning Adaptation models Sentiment analysis Electric potential Profitability Decision making Reinforcement learning Market research Cryptocurrency Telecommunication computing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：